Pluribus Networks, which over the past 10-plus years has built out its software-defined networking (SDN) portfolio to compete with such established vendors as Cisco Systems and Juniper Networks, just got another $20 million injection in its latest funding round.
Leading the funding, which was announced this week, was Morgan Stanley Expansion Capital. Pluribus will use the money to support R&D efforts into multiple new product projects and hire more sales and marketing people as the company works to extend its reach into the burgeoning distributed cloud market and adjacent sectors.
The new funding round brings the total amount raised by Pluribus, which has about 120 employees and 30 full-time contractors, to $145 million.
“To be a serious player in the networking space requires capital,” Mike Capuano, Pluribus’ chief marketing officer, told Enterprise Networking Planet. “We have software and hardware and we’ve seen good acceleration, 50 percent year-over-year growth [and] that’s what we’re intending for next year as well. We just need more sales and marketing resources to support that, as well as with a growing customer base, we’ve seen new opportunities for new products and adjacent markets. That’s also what we want to use the money for, to fuel R&D to allow us to pursue a few new product initiatives in some adjacent markets.”
An SDN Player
Pluribus has been in the SDN space for years, offering enterprises its software solutions, the Linux-based Netvisor ONE operating system and virtualization software and its Adaptive Cloud Fabric, which delivers such features as automation, visibility into network traffic and network segmentation for distributed network architectures.
The combination of products brings cloud-like capabilities, such as scale, elasticity and adaptability, to enterprise data centers as well as service providers, including cloud providers, telecommunications firms, and mobile network operators.
“The fabric can be inside a single data center across some number of racks where it can actually stretch very easily across multiple sites,” Capuano said. “That’s the most typical use case for us and it allows enterprise customers or service providers to basically cloudify their data center network and they can deploy a service in seconds — one command, two commands. That service can be deployed with consistency across the entire fabric, across every switch.”
Software is Hardware-Agnostic
The software can run on Pluribus’ own Freedom series data center switches — which offer 10 Gigabit Ethernet, 25GbE, and 100GbE options and will soon include a 400GbE appliance — or on white box systems from such partners as Dell Technologies and Edgecore.
The vendor also supports other approaches, like Border Gateway Protocol (BGP) VPN, which is a protocol-based approach that enables network operators to offer VPN service. He described it as “more of an imperative type of approach, where you have to configure each box with the protocol. We do provide that choice to customers. Most customers choose to use the SDN functionality because it’s so much easier.”
SDN hit the industry almost a decade ago, part of a larger trend to make data center infrastructure, including compute and storage, more automated and flexible and reduce expenses. The future of SDN as a technology term and trend has been debated in recent years. Gartner analysts in 2019 said in its Hype Cycle report for networking that for the most part, SDN had become obsolete, though the software defined-WAN (SD-WAN) technology that spun out of the movement has become mainstream.
However, others are seeing a robust market. A report on MarketsandMarkets predicted the global SDN market will grow from $13.7 billion last year to $32.7 billion by 2025, driven in part by growing investments by cloud service providers to automate their network infrastructures and increasing demand for cloud services.
Meeting the Demand
Pluribus is aiming to address those demands.
“Anybody who deploys Pluribus — Netvisor ONE and the Adaptive Cloud Fabric — gets full visibility of every flow going across the fabric, every endpoint, every port, with no external equipment,” Capuano said. “You don’t need any taps or probes or packet rovers. The network fabric takes care of all that. That’s the level of visibility. What our customers are asking for is simplicity, visibility and the ability to move at cloud speed. They want to transform that on-prem network.”
Another draw for enterprises is that many are building active-active data centers where the facilities all are active and can support an application at any time, which are becoming key in an increasingly distributed IT environment. These enterprises are looking for availability zone similar to those in public clouds and want to reduce their recovery time objective (RTO) and recovery point objective (RPO) in disaster recovery situations from a day or more down to minutes, he said.
Service providers are focused on becoming more distributed, deploying compute capabilities outside of the central data center and into more regional data centers. Pluribus offers a data center switch specifically designed for such environments.
“That’s where we see the telcos with the mobile network operators,” Capuano said. “We’re built in as the fabric of their virtualized 4G and 5G mobile cores. We also do a lot of business with cloud service providers who often have distributed locations. They want to place workloads close to the end user or to help the end user also connect into the public cloud.”
Capuano said Pluribus officials know the company is swimming in a pool that includes such large rivals as Cisco, Juniper, and Arista Networks. However, those vendors bring traditional legacy approaches in terms of such tasks as configuration. They also tend to bring in external automation tools, which can be expensive and out of sync with an enterprise’s underlying solutions.
He said the technology closest to what Pluribus offers is Cisco’s Application Centric Infrastructure (ACI) platform, which enables organizations to deploy a service across their fabric. However, a key difference is that unlike Cisco, Pluribus doesn’t use controllers, which can add to cost and complexity.
“What we do is leverage the multi-core CPUs in the switches themselves as the compute substrate, and that’s how we run our SDN deployments,” Capuano said. “We’ve just written some clever code and integrated this automation and SDN control plane into the operating system, which has huge cost benefits [in terms of] space power simplicity. But it also has this nice benefit of no integration.”
Every feature Pluribus adds is automated and tested. In network automation, “it’s a pretty heavy lift if you’re doing it through scripting and Python and Ansible and all that,” he said. “For us, it’s just built in and it’s always tested and it always works right out of the box.”
Read next: Best DevOps Tools & Software of 2021